SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To explore the utility of the novel iron indices hepcidin, reticulocyte hemoglobin content (Ret-Hgb), and erythrocyte (red blood cell) hemoglobin content (RBC-Hgb) for detection of iron deficiency in rheumatoid arthritis (RA) patients with anemia and active inflammation and to compare these indices with conventional parameters of iron deficiency.

Methods

Blood samples from 106 outpatients with RA were analyzed in a cross-sectional exploratory study. Forty patients were classified as having either iron deficiency anemia (IDA), anemia of chronic disease (ACD), their combination (IDA/ACD), or “other anemia” based on biochemical parameters for inflammation and iron deficiency. The ability of serum and urine hepcidin, Ret-Hgb, and RBC-Hgb measurement to discriminate among these states was evaluated.

Results

Hepcidin content in serum from patients in the IDA group as well as that from patients in the combined IDA/ACD group differed significantly from that in serum from patients in the ACD group. This difference was also observed with hepcidin in urine, Ret-Hgb, and RBC-Hgb, although with less significance. The area under the receiver operating characteristic curve for serum hepcidin was 0.88 for the comparison of IDA/ACD patients with ACD patients and 0.92 for the comparison of the combined IDA group and IDA/ACD group to all other patients with anemia. Hepcidin at <2.4 nmoles/liter had a sensitivity of 89% and a specificity of 88% to distinguish IDA/ACD from ACD. Both Ret-Hgb and RBC-Hgb measurements also allowed differentiation between these latter groups, with a sensitivity of 67% and 89%, respectively, and a specificity of 100% and 75%, respectively.

Conclusion

Serum hepcidin and, to a lesser extent, urine hepcidin, Ret-Hgb, and RBC-Hgb, are potential useful indicators for detecting iron deficiency in RA patients with anemia and active inflammation.

Rheumatoid anemia is typically multifactorial and related to the chronic inflammatory nature of the disease, and includes iron deficiency, mainly due to therapy-induced gastrointestinal blood loss. The detection of iron deficiency in a population with anemia of chronic disease (ACD) is of clinical relevance, since 1) iron deficiency anemia (IDA) is treatable, 2) the diagnosis may prelude the need for further investigations into the cause of the anemia, and 3) it may prevent unnecessary prescription of iron supplementation.

Unfortunately, diagnosis of iron deficiency with currently available laboratory parameters is hampered by the lack of a gold standard and is even more complex and nonspecific when concomitant inflammatory conditions are present (1, 2). Proinflammatory stimuli contribute to ACD directly by inhibition of erythropoiesis and indirectly by decreasing the iron available for heme synthesis (3). The latter may be attributed to inflammation-induced increased levels of the iron regulatory peptide hepcidin. Elevated hepcidin levels reduce intestinal iron absorption as well as iron release from macrophages through interaction, internalization, and degradation of the cellular iron exporter ferroportin, resulting in iron sequestration in the reticuloendothelial system (4, 5). Consequently, the total body iron content is normal, but less iron is available for erythropoiesis. In contrast, in IDA, in which there is an absolute iron deficiency, hepcidin is suppressed, which leads to induction of iron absorption from the gut. Since hepcidin has been shown to be differently affected by inflammation and iron deficiency, it has been advocated as a potential biomarker to assess iron deficiency in patients with inflammatory conditions (6–9).

The hemoglobin (Hgb) contents of erythrocytes (red blood cells [RBCs]) and of reticulocytes (RBC-Hgb and Ret-Hgb) have also been explored as biomarkers for diagnosing iron deficiency (10–13). Ret-Hgb and RBC-Hgb reflect the hemoglobinization of the earliest-released reticulocytes, which circulate for only 1–2 days, and that of mature erythrocytes, respectively (14). Reduced Ret-Hgb and RBC-Hgb values indicate that the iron supply for the bone marrow is too low to allow normal hemoglobinization (12). Since these parameters are not directly affected by inflammation but are functional measures of iron availability, they might be of relevance to detect iron deficiency in conditions of chronic inflammation, such as rheumatoid arthritis (RA) (11). To assess the added diagnostic value of measuring serum and urine hepcidin, Ret-Hgb, and RBC-Hgb (measured as reticulocyte Hgb equivalents [Ret-He] and RBC Hgb equivalents [RBC-He], respectively), for the detection of iron deficiency in patients with anemia and inflammation, we explored their diagnostic characteristics and compared them with conventional indices of iron status in a cross-sectional study of 106 outpatients with RA.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Study population.

The study population consisted of 106 patients with RA who consecutively attended the outpatient clinic at Radboud University Nijmegen Medical Centre, a tertiary care facility, for routine evaluation between March 2007 and March 2008. In this exploratory cross-sectional study, no other inclusion or exclusion criteria were applied. The local ethics committee approved the study, and written informed consent was obtained from all participants, in compliance with the Declaration of Helsinki.

Anemia was defined, according to the World Health Organization criteria, as a hemoglobin concentration of <8.1 mmole/liter (13 gm/dl) for men and <7.4 mmoles/liter (12 gm/dl) for women. Patients with anemia were divided into 4 groups, based on biochemical parameters for inflammation and iron deficiency (see below).

Laboratory analysis.

Biochemical marker levels and hematologic indices were determined in all participants on the day of their routine outpatient visit. Blood counts were measured with an automated hematology analyzer (Sysmex 2100; Sysmex). Hemoglobin content parameters were provided as Ret-He and RBC-He and are reported in fmoles (1 fmole = 16.11 pg). The following parameters were measured using an Abbott Aeroset analyzer: iron, plasma transferrin concentration, total iron-binding capacity (TIBC), and creatinine (reagents from Roche Diagnostics), C-reactive protein (CRP), alanine aminotransferase, lactate dehydrogenase, and haptoglobin (reagents from Abbott Laboratories), and bilirubin (reagent produced in-house). The percent plasma transferrin saturation was calculated as 100 times the serum iron concentration divided by the TIBC. Ferritin, vitamin B12, and folic acid were quantified with a solid-phase 2-site chemiluminescent immunometric assay (Immulite 2500; Diagnostic Products). The erythrocyte sedimentation rate (ESR) was determined using a StaRRsed Compact ESR analyzer (Mechatronics). Serum concentrations of soluble transferrin receptor (sTfR) were measured immunonephelometrically on a BN II System (Dade-Behring). The sTfR index (sTfR divided by log-transformed ferritin values) was calculated (15). Thalassemia was discriminated from IDA using previously described formulas (14, 16). The glomerular filtration rate was assessed using the Modification of Diet in Renal Disease formula (17).

Serum hepcidin-25 was measured by a combination of weak cation exchange chromatography and time-of-flight mass spectrometry, as described previously (9). Serum hepcidin concentrations were expressed as nmoles/liter (1 nmole/liter = 2.789 μg/ml). The lower limit of detection with this method was 0.5 nmoles/liter; average coefficients of variation were 2.7% (intra-run) and 6.5% (inter-run). The median reference level of serum hepcidin is 4.2 nmoles/liter, with a range of 0.5–13.9 nmoles/liter (18). In this study, serum hepcidin concentrations below the lower limit of detection were recorded as 0.4 nmoles/liter for statistical calculations.

Urine hepcidin was assessed by a method similar to that used for serum hepcidin (19). Values were normalized to urine creatinine values and reported in nmoles/mmole creatinine. The lower limit of detection for urine hepcidin was 0.05 nmoles/nmole creatinine.

Classification of anemia.

In the RA patients with anemia, we classified the anemia as IDA, ACD, or a combination of the two (IDA/ACD) by modifying previously described algorithms (3, 6, 20) to a scheme appropriate to our patient population. Specifically, patients with anemia as defined above were classified as having IDA if active inflammation (defined as a CRP level of ≥10 mg/ml or an ESR of ≥30 mm/hour) was absent and at least 1 of the following 2 conditions was met: 1) transferrin saturation <20% and ferritin level <30 μg/ml; 2) sTfR index ≥1 mg/μg. Patients were classified as having ACD if active inflammation was present and at least 1 of the following 2 conditions was met: 1) transferrin saturation <20% and ferritin level ≥100 μg/ml; 2) sTfR index <1 mg/μg and ferritin level ≥30 μg/ml. Patients were classified as having IDA/ACD if active inflammation was present and at least 1 of the following 2 conditions was met: 1) transferrin saturation <20% and ferritin level <100 μg/ml; 2) sTfR index ≥1 mg/μg. Patients with anemia that could not be classified according to these definitions were categorized as having “other anemia.” RA patients without anemia were studied as controls.

Assessment of disease activity.

RA disease activity was assessed using the Disease Activity Score in 28 joints (DAS28) (21). A DAS28 score of ≤3.2 is considered to represent low disease activity (22).

Statistical analysis.

The median value and interquartile range (IQR) for each continuous parameter were determined. Depending on the data distribution (as assessed by Kolmogorov-Smirnov and Levene's normality testing), standard parametric and nonparametric tests were applied. Spearman's rank correlation testing was used for correlation analysis. Receiver operating characteristic (ROC) analysis was performed to evaluate the accuracy with which hepcidin and the hemoglobin content parameters differentiated the groups of patients with anemia. The area under the ROC curve (AUC) was generated, including the sensitivity and specificity for each possible cutoff of the specific parameter. An AUC value of 1 represents the perfect discriminating test. The Youden index ([sensitivity + specificity] − 1) was used to select the optimal cutoff value. P values less than 0.05 were considered significant. All statistical analyses were carried out using SPSS version 16.0.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Population characteristics.

The characteristics of the 106 RA patients are shown in Table 1. Forty patients (37.7%) were anemic; of these 40, 10 (25.0%), 9 (22.5%), 8 (20.0%), and 13 (32.5%) fulfilled the predefined criteria for IDA, IDA/ACD, ACD, and other anemia, respectively. Renal function, results of liver biochemistry tests, and vitamin B12 and folic acid levels were comparable among groups, and none of the patients had signs of hemolysis or thalassemia. Disease activity as assessed by the DAS28 was significantly increased in anemic versus non-anemic patients; however, the score was similar among the different anemia groups.

Table 1. Baseline demographic characteristics and laboratory parameters in the RA patients studied*
 No anemia (n = 66)IDA (n = 10)IDA/ACD (n = 9)ACD (n = 8)Other anemia (n = 13)P
  • *

    Except where indicated otherwise, values are the median (interquartile range). RA = rheumatoid arthritis; IDA = iron deficiency anemia; ACD = anemia of chronic disease; DAS28 = Disease Activity Score in 28 joints; MCV = mean corpuscular volume; TSAT = transferrin saturation; sTfR = soluble transferrin receptor; sTfR index = sTfR/log ferritin; MDRD = Modification of Diet in Renal Disease; ALT = alanine aminotransferase; LDH = lactate dehydrogenase; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate.

  • Comparison between all groups, by analysis of variance, Kruskal-Wallis test, or chi-square test, as appropriate.

  • P ≤ 0.01 versus ACD group, by independent t-test, Mann-Whitney test, or chi-square test, as appropriate.

  • §

    P < 0.05 versus ACD group, by independent t-test, Mann-Whitney test, or chi-square test, as appropriate.

  • P ≤ 0.0001 versus ACD group, by independent t-test, Mann-Whitney test, or chi-square test, as appropriate.

Female, no. (%)45 (69.3)8 (80.0)4 (44.4)7 (87.5)7 (53.8)0.25
Age, years61.0 (49.5–68.0)50.5 (37.5–67.8)§73.0 (57.0–79.5)68.5 (66.0–74.0)62.0 (51.5–74.5)0.03
DAS28 score3.1 (2.3–3.9)3.7 (2.3–5.3)4.2 (2.7–5.4)4.1 (3.9–4.7)3.9 (3.1–4.3)0.02
DAS28 ≤3.2, no. (%)33 (50.0)5 (50.0)§3 (33.3)0 (0)3 (42.9)0.04
Hemoglobin, mmoles/liter8.2 (7.8–8.6)6.8 (6.4–7.1)7.0 (5.9–7.2)7.0 (6.6–7.3)6.9 (6.6–7.2)<0.001
MCV, fl88.1 (84.3–92.0)82.9 (80.8–87.3)§86.9 (78.2–91.1)91.4 (89.9–96.6)89.4 (85.7–94.4)0.03
Iron, μmoles/liter14.0 (10.0–18.0)6.0 (6.0–15.0)§6.0 (4.5–8.5)10.5 (8.0–13.8)12.0 (8.0–15.5)0.001
Ferritin, μg/liter90.0 (47.8–141)11.5 (8.3–22.8)53.0 (11.0–64.5)191 (102–262)84.0 (61.0–133)<0.001
Transferrin, gm/liter2.4 (2.2–2.7)2.7 (2.4–3.0)2.5 (2.2–2.7)§2.1 (1.8–2.4)2.2 (2.1–2.3)0.06
TSAT, %22.4 (16.8–28.1)9.7 (7.5–22.8)9.7 (7.2–13.6)21.9 (17.9–26.5)21.1 (14.0–27.9)0.001
sTfR, mg/liter1.5 (1.2–1.8)2.5 (2.0–3.1)2.4 (1.9–2.8)1.4 (1.3–1.9)1.6 (1.2–1.9)<0.001
sTfR index, mg/μg0.7 (0.6–1.0)2.5 (2.5–3.1)1.4 (1.2–2.3)0.6 (0.6–0.9)0.8 (0.7–0.9)<0.001
MDRD, ml/minute/m297.6 (76.9–110)102 (79.5–113)101 (72.1–112)101 (84.1–113)88.3 (64.5–109)0.78
ALT, units/liter22.0 (17.0–8.5)21.5 (14.8–32.0)16.0 (12.5–21.0)21.5 (15.3–27.3)20.0 (16.0–31.0)0.36
LDH, units/liter400 (361–473)447 (358–479)393 (331–473)393 (342–475)454 (390–521)0.52
Haptoglobin, gm/liter1.4 (0.9–2.0)1.1 (0.8–1.9)1.7 (1.4–2.4)2.0 (1.5–3.1)1.6 (0.9–2.0)0.09
Folic acid, nmoles/liter25.9 (15.5–36.6)20.3 (14.2–38.5)33.8 (16.5–46.5)24.4 (16.9–38.6)24.9 (16.2–42.7)0.90
Vitamin B12, pmoles/liter267 (196–378)425 (212–601)301 (214–629)281 (176–485)280 (157–406)0.39
CRP, mg/liter4.0 (4.0–10.3)4.0 (4.0–7.0)13.0 (9.0–19.0)17.0 (10.5–27.5)13.0 (5.5–27.5)<0.001
ESR, mm/hour11.0 (5.0–20.3)15.5 (8.0–20.5)32.0 (26.0–52.0)46.0 (26.3–54.8)16.0 (10.0–35.5)<0.001

In accordance with the selection criteria, median CRP and ESR values were significantly increased in both the IDA/ACD and the ACD groups. Notably, 20 (30.3%) of the non-anemic RA patients and 7 (53.8%) of the patients in the other anemia category also had increased levels of inflammation markers. Further group differences apart from those due to selection criteria were as follows: IDA patients were younger (median age 50.5 years, versus 73.0, 68.5, and 62.0 years in the IDA/ACD, ACD, and other anemia groups, respectively) and had significantly lower mean corpuscular volume (MCV) compared to the patients with ACD. MCV values in the other anemia groups were similar to those in the non-anemic controls.

Hepcidin and hemoglobin content.

Median serum hepcidin concentrations in both the IDA and the IDA/ACD groups were significantly different from that in the ACD group (median [IQR] nmoles/liter 7.4 [2.6–11.0], 0.4 [0.4–0.8], and 0.7 [0.4–2.2] in the ACD, IDA, and IDA/ACD groups, respectively [P = 0.001 ACD versus IDA, P = 0.009 ACD versus IDA/ACD]) (Figure 1 and Table 2). This difference was also demonstrated in the comparison of IDA patients or IDA/ACD patients with non-anemic controls (P < 0.001 and P = 0.001, respectively). In contrast, serum hepcidin concentrations in ACD patients were similar to those in controls (P = 0.20). A similar trend was observed with regard to urine hepcidin, although statistical significance between anemia groups was only reached in the comparison between IDA and ACD patients (P = 0.008). In addition, ACD patients differed from IDA patients by significantly higher RBC-Hgb values (P = 0.03) and from IDA/ACD patients by significantly higher Ret-Hgb values (P = 0.02). None of the hemoglobin content parameters differed between the ACD group and the non-anemic controls.

thumbnail image

Figure 1. Distribution of markers of hepcidin and hemoglobin content in the predefined groups of rheumatoid arthritis patients. Data are presented as box plots, where the boxes represent the 25th to 75th percentiles, the lines within the boxes represent the median, and the lines outside the boxes represent the minimal and maximal values. The y-axes for serum and urine hepcidin levels are log transformed. ∗ = P < 0.05; ∗∗ = P ≤ 0.01; ∗∗∗ = P ≤ 0.001, by independent t-test or nonparametric Mann-Whitney test. IDA = iron deficiency anemia; ACD = anemia of chronic disease; IDA/ACD = combined IDA and ACD; Cr = creatinine; Ret-Hgb = reticulocyte hemoglobin content; RBC-Hgb = erythrocyte (red blood cell) hemoglobin content.

Download figure to PowerPoint

Table 2. Levels of hepcidin and novel indices of hemoglobin content in RA patients with various forms of anemia and those without anemia*
 No anemia (n = 66)IDA (n = 10)IDA/ACD (n = 9)ACD (n = 8)Other anemia (n = 13)P
  • *

    Values are the median (interquartile range). RA = rheumatoid arthritis; IDA = iron deficiency anemia; ACD = anemia of chronic disease; Ret-Hgb = reticulocyte hemoglobin content; RBC-Hgb = erythrocyte (red blood cell) hemoglobin content.

  • Comparison between all groups, by analysis of variance or Kruskal-Wallis test, as appropriate.

  • P ≤ 0.001 versus ACD group, by independent t-test or Mann-Whitney test, as appropriate.

  • §

    P ≤ 0.01 versus ACD group, by independent t-test or Mann-Whitney test, as appropriate.

  • P < 0.05 versus ACD group, by independent t-test or Mann-Whitney test, as appropriate.

Serum hepcidin, nmoles/liter3.5 (2.8–6.2)0.4 (0.4–0.8)0.7 (0.4–2.2)§7.4 (2.6–11.0)4.6 (2.6–5.9)<0.001
Urine hepcidin, nmoles/mmole Cr0.6 (0.3–1.4)0.06 (0.02–0.2)§0.2 (0.04–1.2)0.9 (0.3–2.4)0.8 (0.4–2.4)0.004
Ret-Hgb, fmoles2.1 (2.0–2.3)1.8 (1.7–2.1)1.8 (1.5–2.0)2.1 (2.0–2.1)2.2 (2.0–2.2)0.006
RBC-Hgb, fmoles1.9 (1.8–2.0)1.7 (1.6–1.8)1.7 (1.5–1.8)1.9 (1.8–2.0)1.9 (1.8–2.0)0.001

Correlation of hepcidin levels with Ret-Hgb and RBC-Hgb levels.

The relationship between novel and conventional parameters of iron in RA patients with anemia were explored. A strong association was observed between serum concentrations and urine concentrations of hepcidin (Spearman's r = 0.84, P < 0.001). Serum hepcidin levels correlated slightly better with ferritin levels and with the sTfR index than did urine hepcidin levels (r = 0.79 and r = 0.79, respectively, versus r = 0.71 and r = 0.68). Ret-Hgb and RBC-Hgb levels correlated strongly with one another (r = 0.88, P < 0.001); however neither was strongly correlated with serum hepcidin levels (r = 0.39, P = 0.014 and r = 0.34, P = 0.035, respectively). Data on these findings in individual patients are available at http://www.hepcidinanalysis.com/documents/SupplementalFigure_vanSantenetal_ArthritisRheum2011.pdf.

Diagnostic characteristics for detection of iron deficiency.

ROC analysis was used to assess the characteristics of serum hepcidin, RBC-Hgb, and Ret-Hgb as diagnostic tests for iron deficiency in the RA patients with anemia in each specific category. Urine hepcidin was not further analyzed because of its inferiority to serum hepcidin for differentiating IDA/ACD from ACD.

First, we performed analyses to detect iron deficiency in the 2 patient groups with inflammation. The AUC for serum hepcidin in the combined IDA/ACD group versus the ACD group was 0.88 (P = 0.009, SE 0.09) (Figure 2). The optimal cutoff value for hepcidin was 2.4 nmoles/liter, as calculated by the Youden index. A hepcidin level lower than this cutoff value distinguished the IDA/ACD patients from the ACD patients with a sensitivity of 89% and a specificity of 88%. The AUC for Ret-Hgb in the combined IDA/ACD patients versus the ACD patients was 0.81 (P = 0.03, SE 0.12) with an optimal cutoff of <1.9 fmoles (sensitivity 67% and specificity 100%), and the AUC for RBC-Hgb was 0.78 (P = 0.05, SE 0.12) with an optimal cutoff of <1.8 fmoles (sensitivity of 89% and specificity of 75%).

thumbnail image

Figure 2. Receiver operator characteristic analysis of hepcidin and hemoglobin content parameters as diagnostic tests for iron deficiency in rheumatoid arthritis patients with specific categories of anemia. Area under the curve (AUC) values are shown. See Figure 1 for other definitions.

Download figure to PowerPoint

Furthermore, the clinical accuracy of hepcidin and iron content indices in the discrimination between patients with iron deficiency anemia (IDA and IDA/ACD [n = 19]) compared to all other patients with anemia (ACD and other anemia [n = 21]) was explored to investigate the suitability of these parameters for the detection of iron deficiency among all of the RA patients with anemia, thus including those without inflammation. The AUC for serum hepcidin in this analysis was 0.92 (P < 0.001, SE 0.05). Application of the same analyses for the hemoglobin content indices resulted in an AUC of 0.75 (P = 0.006, SE 0.08) and 0.80 (P = 0.001, SE 0.07) for Ret-Hgb and RBC-Hgb, respectively. These results indicate that hepcidin is superior to Ret-Hgb and RBC-Hgb as a biomarker for the detection of iron deficiency among RA patients with anemia.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

In this cross-sectional exploratory study of RA patients attending an outpatient clinic, serum and urine hepcidin levels and hemoglobin content parameters were assessed in an effort to identify markers of iron deficiency anemia in patients with inflammatory disease. The detection of absolute iron deficiency in RA patients with anemia is a clinical challenge, since conventional iron status screening parameters, e.g., transferrin, iron, and ferritin, are influenced by acute-phase responses (23). TfR reflects iron stores linearly to the degree of iron demand of erythroblasts, but inflammatory stimuli may suppress its concentrations by inhibition of erythropoietic activity (2, 24). The sTfR index was developed to overcome this effect (15) and has gained merit in diagnostic algorithms together with Ret-Hgb (1, 8, 12); however, each of these markers has its own disadvantages in predicting iron deficiency (2, 23–25). Figure 3 provides an overview of serum iron parameters and hemoglobin content parameters in patients with RA and anemia.

thumbnail image

Figure 3. Overview of indicators of serum iron status in rheumatoid arthritis patients with ACD and in those with IDA. A, ACD. In states of inflammation, increased hepcidin quantities are released into the circulation by hepatocytes (36, 37). Hepcidin binds, internalizes, and degrades the iron exporter ferroportin (FPN), which results in a decrease in iron export from enterocytes and macrophages into the bloodstream (38). As a consequence, serum iron and transferrin saturation (TSAT) decrease while iron stored within macrophages increases, leading to elevated serum ferritin levels. The inflammation-induced suppression of erythropoiesis counteracts the decreased availability of iron (3, 24).Therefore, the serum (or soluble) transferrin receptor (sTfR) concentration and the hemoglobin content of reticulocytes and erythrocytes may remain relatively normal (3, 24, 33, 39) (see also Figure 1). B, IDA. Iron deficiency reduces hepcidin (even in the presence of inflammation [7]), which results in maximal release of iron from enterocytes and macrophages into the circulation. Nevertheless, ferritin and iron levels and TSAT decrease. Consequently, the limited iron supply to the bone marrow will increase the sTfR concentration to allow maximal iron uptake by erythroblasts. Ret-Hgb and RBC-Hgb levels are reduced, indicating insufficient iron availability for erythropoiesis (10, 31). RES = reticuloendothelial system (see Figure 1 for other definitions).

Download figure to PowerPoint

In our RA patients, serum hepcidin levels correlated highly with urine hepcidin and ferritin levels and the sTfR index, but not with hemoglobin content parameters. Moreover, in the comparison of categories of anemia as defined by conventional parameters, serum hepcidin performed well as a diagnostic test of iron deficiency, including when inflammation was present, and discriminated both the IDA group and the combined IDA/ACD group from patients with ACD. Urine hepcidin, RBC-Hgb, and Ret-Hgb also enabled differentiation between iron deficiency and ACD, albeit with lower statistical significance compared to serum hepcidin.

Hepcidin levels below 2.4 nmoles/liter discriminated IDA and IDA/ACD patients from ACD patients with RA. Apparently, the induction of hepcidin by inflammation is blunted by low iron stores. This suppression of hepcidin in anemic patients with iron deficiency and with inflammation is corroborated by findings of animal studies (26) and in various other patient groups, i.e., in a mixed population of patients with chronic infection, autoimmune disease, or malignancy (20, 27), in cancer patients with inflammation (6), and in patients with acute inflammation in an intensive care unit (28). In contrast, in another heterogeneous population of patients with chronic inflammation and anemia, hepcidin levels did not differentiate patients with IDA/ACD from those with ACD (8); this may be attributed to the use of a different stratification scheme than was used in the present study.

Since hepcidin concentrations at the lowest level of detection (<0.5 nmoles/liter) were indicative of IDA, we explored the added value of urine hepcidin as a more sensitive marker in the lower concentration range. Although measuring urine hepcidin instead of serum hepcidin did not add to the level of diagnostic accuracy in the current and previous studies (18), it might be an attractive alternative in the developing world, as a noninvasive tool for assessing iron deficiency as a safety measure before starting oral iron administration programs (29).

Our findings enable us to propose a novel model for selecting RA patients who would likely benefit from iron supplementation (Figure 4). Since rheumatoid anemia is heterogeneous, more than one single parameter might be necessary to allow diagnosis of iron deficiency. We reasoned that patients with inflammation (test 1) and a hepcidin level of <2.4 nmoles/liter (test 2) will have iron deficiency, while hepcidin levels ≥7.6 nmoles/liter are diagnostic for ACD. Patients whose hepcidin levels fell between 2.4 and 7.6 nmoles/liter were further characterized using Ret-Hgb (test 3), with levels of <1.9 fmoles being indicative of concomitant iron deficiency. For IDA, a hepcidin level of <0.5 nmoles/liter was highly specific. Using these definitions to assess the need for iron supplementation, all patients from the combined IDA/ACD group would benefit from such supplements, with no ACD patients receiving this treatment unnecessarily. This model needs to be confirmed in other populations and may benefit from further international standardization of hepcidin measurements (30).

thumbnail image

Figure 4. Proposed algorithm for assessing which rheumatoid arthritis patients with anemia may benefit from iron treatment, based on test characteristics described in the text. CRP = C-reactive protein; ESR = erythrocyte sedimentation rate (see Figure 1 for other definitions).

Download figure to PowerPoint

The study also provides insights regarding hepcidin and hemoglobin content levels in IDA and ACD patients in comparison to control RA patients without anemia. Both Ret-Hgb and RBC-Hgb discriminated IDA patients from non-anemic RA controls. These observations are consistent with previous reports of convincing ROC characteristics demonstrating the ability of these indices to detect iron deficiency in otherwise healthy subjects and patients undergoing dialysis (10, 14, 31, 32). Interestingly, in RA patients with ACD, we found that levels of hepcidin and the hemoglobin content parameters did not differ significantly from those obtained in non-anemic patients. It appears that in RA patients with chronic disease activity, hepcidin concentrations do not increase to levels seen in acute inflammation. In accordance with the current concept of a causal role of hepcidin in iron-restricted erythropoiesis, the absence of significantly increased hepcidin levels in the ACD group likely explains the similarities in hemoglobin content parameters between the ACD and the non-anemic control RA patients. Moreover, the RA control group differs from a healthy population in terms of, e.g., the presence of inflammatory markers and/or other indices of disease activity.

There are very few reported studies comparing hemoglobin content parameters in healthy controls with those in patients with ACD. In a heterogeneous cohort of ACD patients, the mean cell hemoglobin level was similar to that in age-matched controls, despite elevated hepcidin levels (27). In another mixed group of patients with ACD, levels of hemoglobin content parameters were normal as compared to reference values (8). A similar study showed a nonsignificant trend toward reduced Ret-Hgb concentrations in patients with inflammation; no hepcidin data were available (33). Interestingly, increased hepcidin levels do not necessarily lead to reduced hemoglobin content, despite the induction of iron withholding by hepcidin. Another mechanism related to ferroportin expression on reticulocytes is probably involved (34, 35). Clearly, more and larger studies are needed to further assess levels of hepcidin in patients with chronic inflammation and its association with parameters of hemoglobin content.

To our knowledge, this is the first study of both hepcidin and hemoglobin content parameters in an RA cohort. In the absence of a gold standard, we were dependent on conventional iron parameters to classify the anemia groups IDA, IDA/ACD, and ACD. Notably, this classification scheme precludes direct comparison of hepcidin or hemoglobin content parameters with any of the conventional parameters used for detection of true iron deficiency in RA patients. We believe the strength of this study lies in its unbiased cross-sectional inclusion of RA patients, and its limitation lies in the small numbers of patients in the various anemia categories.

What lessons can be learned from this study? Is it time for the rheumatologist to add hepcidin and hemoglobin content parameters to the clinical measurements that are performed? The present results clearly demonstrate the potential utility of serum hepcidin, and to lesser extent Ret-Hgb and RBC-Hgb, in the detection of iron deficiency in RA patients with inflammation and anemia. Our findings indicate that these novel parameters can provide added value in diagnosing iron deficiency in patients with rheumatoid anemia; larger studies are needed to confirm this.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Swinkels had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Van Santen, van Dongen-Lases, de Vegt, van Riel, van Ede, Swinkels.

Acquisition of data. Van Dongen-Lases, Laarakkers, van Riel, van Ede, Swinkels.

Analysis and interpretation of data. Van Santen, van Dongen-Lases, de Vegt, Laarakkers, van Ede, Swinkels.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We would like to thank Eline Liesting, Liesbeth Peels-van der Snoek, Judith Kulk, and Siem Klaver for help with acquisition of data, Anouk Linssen for assistance in the analysis of the data, and Erwin Wiegerinck for performing serum hepcidin measurements. We are also grateful to J. Harbers for his contribution to the design of the study and to obtaining ethics permission.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES